Large eddy simulations model river vegetation

Vegetation in river systems can have a significant influence on the hydraulics of the river channel, just as the river flow can shape the type and spread of the vegetation. Understanding that interplay is useful for engineering river design and restoration. Some studies have independently investigated the density and streamlining angle of vegetation, but no work had been done to understand the three-dimensional forces and wakes that underlie riverine ecological processes and sediment transport.

Zhang et al. used a large-eddy simulation to model the flow of water past an infinite patch of vegetation modeled by an array of cylinders. They found that the density of the cylinders and their inclination angle were varied, matching how vegetation stems often tilt with the flow of water. They also measured the resulting forces, torques, pressures, velocity fields and shedding frequencies for each vegetation setup, which were validated against experiments to check for numerical accuracy.

The team’s results showed that higher density consistently increases drag, torque and flow irregularity, particularly near the base. This information can help predict channel resistance and bed scour in vegetated areas. They also found the vegetation stem angle had nonlinear effect with larger angles intensifying three-dimensional flow structures.

“Together, this study provides the first map linking angle and density to 3D flow structures in rigid vegetation,” said author Yi Zhang. “It improves engineering predictions of channel resistance, bed shear, and potential scour in vegetated rivers, and supports nature-based solutions for flood mitigation. More broadly, it offers a physics-based tool to connect vegetation posture and packing with river morphodynamics and resilience.”

Source: “Hydrodynamics of inclined cylinder arrays: Effects of streamlining angle and vegetation density,” by Yi Zhang, Subhasish Dey, Jihao Jian, Wei Liu, Xiong Tang, and Siming He, Physics of Fluids (2025). The article can be accessed at https://doi.org/10.1063/5.0289562 .